消防科学与技术 ›› 0, Vol. ›› Issue (): 1472-1476.
• • 上一篇 下一篇
张 毅
出版日期:
发布日期:
作者简介:
Zhang Yi
Online:
Published:
摘要: 摘 要:基于当下灭火救援领域中的突发事件,提出灭火救援专业知识智能匹配算法,该算法基于自然语言处理和注意力机制计算案件描述和消防预警信息之间的语义关系,从而实现相关灭火救援专业知识的匹配。首先基于自然语言处理的方法学习句粒度级别的语义信息,然后基于注意力机制学习词粒度级别的语义信息,最后基于两个级别语义信息的交互,根据信息之间局部差异推断两个句子之间的关系。试验结果表明:该算法具有优异的性能,能够同时从词和句两个粒度上更准确地理解句子,实现基于案件描述的灭火救援专业知识智能匹配。
关键词: 关键词:灭火救援, 智能匹配, 自然语言处理, 注意力机制, 语义信息
Abstract: Abstract: The FDS was used to simulate the lateral point smoke exhaust system in a double-deck shield road tunnel, and the influence of smoke outlet area, space, smoke outlet opening scheme and longitudinal ventilation on the smoke exhaust system is studied. The results show that without longitudinal ventilation, the area within 3 m2 to 5 m2 and space between 60 m and 100 m of smoke outlets have little effect on the effectiveness of smoke exhaust after the fire has stabilized. As the longitudinal ventilation velocity increases and more upstream smoke outlets opens, the smoke exhaust efficiency decreases significantly. The smoke exhaust efficiency for the upper and lower exhaust outlets of the double-deck tunnel is basically the same, and the total smoke exhaust efficiency of the lower tunnel is slightly higher than that of the upper tunnel. It is proposed that in the event of a 20 MW fire in this double-decker tunnel, smoke exhaust is most effective when the longitudinal ventilation velocity is 2 m/s, the smoke outlet space is 60 m, the smoke outlet area is 4 m2, and two upstream and four downstream smoke outlets are opened.
Key words: Key words: double-deck shield tunnel, lateral point smoke exhaust, numerical simulation, smoke exhaust efficiency
张 毅. 灭火救援专业知识智能匹配算法[J]. 消防科学与技术, 0, (): 1472-1476.
Zhang Yi. Intelligent matching algorithm for firefighting and rescue professional knowledge[J]. Fire Science and Technology, 0, (): 1472-1476.
0 / / 推荐
导出引用管理器 EndNote|Ris|BibTeX
链接本文: https://www.xfkj.com.cn/CN/